Plunging Flow Depth Estimation in a Stratified Dam Reservoir Using Neuro-Fuzzy Technique

The cold river water inflow often plunges below the ambient dam reservoir water and becomes density underflow through the reservoir. The hydrodynamics of density currents and plunging are difficult to study in the natural environment and laboratory condition due to small-scale, entrainment and turbulent flows. Numerical modeling of plunging flow and defining of the plunging depth can provide valuable insights for the dam reservoir sedimentation and water quality problem. In this study, an adaptive neuro-fuzzy (NF) approach is proposed to estimate plunging flow depth in dam reservoir. The results of the NF model are compared with two-dimensional hydrodynamic model, artificial neural network (ANN), and multi linear regression (MLR) model results. The two-dimensional model is adapted to simulate density plunging flow simulation through a reservoir with sloping bottom. The model is developed using nonlinear and unsteady continuity, momentum, energy and k-ε turbulence model equations in the Cartesian coordinates. Density flow parameters such as velocity, plunging points, and plunging depths are determined from the simulation and model results. Mean square errors (MSE), mean absolute errors (MAE) and determination coefficient (R2) statistics are used as comparing criteria for the evaluation of the models’ performances. The NF model approach for the data yields the small MSE (1.18 cm), MAE (0.86 cm), and high determination coefficient (0.95–0.98). Based on the comparisons, it was found that the NF computing technique performs better than the other models in plunging flow depth estimation for the particular data sets used in this study.